Tuning the Structure and Performance of Bulk and Porous Vapor Sensors Based on Co-continuous Carbon Nanotube-Filled Blends of Poly(vinylidene fluoride) and Polycarbonates by Varying Melt Viscosity.
Yilong LiYanjun ZhengJürgen PionteckPetra PötschkeBrigitte VoitPublished in: ACS applied materials & interfaces (2020)
This work describes a new concept of porous vapor sensor materials based on co-continuous polycarbonate/poly(vinylidene fluoride)/multiwalled carbon nanotube (PC/PVDF/MWCNT) blend composites. The blend composites were fabricated by melt mixing in a one-step mixing process, and the MWCNT containing component (here PC) was extracted, leaving a MWCNT network on the continuous surface of the remaining component (here PVDF). First, by selecting three PCs with different molecular weights, the blend viscosity ratio and blend fineness and interfacial area were varied. At the chosen blend composition of 40/60 wt %, the desired co-continuous structure was achieved with MWCNTs selectively localized in PC. The conductive polymer composites (CPCs) with low-viscosity PC had the highest conductivity due to a combination of the best MWCNT dispersion and the coarsest blend morphology. The vapor sensing of CPC sensor materials with 1 wt % MWCNT was tested using saturated vapors of dichloromethane, acetone, tetrahydrofuran, and ethyl acetate, showing good interaction with PC. The compact compression molded CPC materials with low-viscosity PC showed the lowest relative resistance changes (Rrel) during the cyclic sensing tests, but a better recovery compared to corresponding CPCs with medium and high viscosity PC. The porous CPC sensors showed remarkable vapor sensing performance compared to the corresponding compact sensors with better sensing stability, reproducibility, and reversibility. Scanning electron microscopy (SEM) confirmed that a fraction of the nanotubes remained on the surface of the continuous, nonsoluble PVDF after PC extraction. The porous sensor material from which the low-viscosity PC was extracted showed the highest Rrel (e.g., around 1300% after 100 s immersion in acetone vapor) compared to all other organic vapors investigated. The difference in vapor measurement between compact and porous sensor materials was attributed to the different sensing mechanisms of polymer swelling for the compact and vapor absorption on the free CNT networks for the porous samples.